[gmx-users] GROMOS (ffG43a2) dihedral potentials

[McMullen, Roger L]

Greetings all.

I am attempting to construct a GROMOS topology file (to be used with
GROMACS) and had a couple of questions about the ffG43a2 bonded parameter
file with respect to the dihedral section.  The following are few lines from
the dihedral section of ffG43a2bon.itp.

1) #define gd_3    180.000   7.11    2
    ;  -C-OA-  (carboxyl)    4.0

2) #define gd_14   0.000     3.77     3
    ;  -C,CHn, SI-NT,NL,OA (sugar)  3

3) #define gd_8     0.000     2.09     2
    ;  O-CH1-CHn-O    0.5

In example 1, is it safe to assume that C and OA are the ith (first) and lth
(fourth) atoms of the four atoms that make up the dihedral angle.  If so,
would example 2 be describing a system in which one could have a combination
of C with NT, NL, or OA; or a combination of CHn with NT, NL, or OA; or a
combination of SI with NT, NL, or OA.  Does the last example (example 3)
describe a system in which all four atoms of the dihedral are being
described in that particular order?  Is this a special case where all atoms
are listed?  Also, in all cases there is a number on the comment line after
the atom types:  Example 1 - 4.0, Example 2 - 3, and Example 3 - 0.5.  What
do these numbers represent?

As always, thanks for your patience and help.

Cheers,
Roger